Skip to content
GLACIS GLACIS Documentation

Sidecar Configuration

This page documents all configuration options for GLACIS sidecars.

Configuration File

Create a glacis.config.ts (or .js) file in your project:

import { defineConfig } from '@glacis/sidecar';


export default defineConfig({
  // Required
  orgId: process.env.GLACIS_ORG_ID!,
  apiKey: process.env.GLACIS_API_KEY!,


  // Provider configuration
  provider: {
    type: 'openai',
    apiKey: process.env.OPENAI_API_KEY!,
    baseUrl: 'https://api.openai.com/v1', // Optional
  },


  // Sampling configuration
  sampling: {
    rate: 100,           // 1 in N requests get L2 attestation
    policies: ['toxicity', 'pii', 'bias'],
    rules: [],           // Adaptive sampling rules
  },


  // Service endpoints
  services: {
    witness: 'https://witness.glacis.io',
    receipt: 'https://receipts.glacis.io',
  },


  // Logging
  logging: {
    level: 'info',       // 'debug' | 'info' | 'warn' | 'error'
    format: 'json',      // 'json' | 'text'
  },
});

Environment Variables

VariableRequiredDescription
GLACIS_ORG_IDYesYour organization ID (org_...)
GLACIS_API_KEYYesYour API key (glc_...)
OPENAI_API_KEYDependsRequired for OpenAI provider
ANTHROPIC_API_KEYDependsRequired for Anthropic provider
SAMPLING_RATENoOverride sampling rate

Provider Configuration

OpenAI

provider: {
  type: 'openai',
  apiKey: process.env.OPENAI_API_KEY!,
  organization: 'org-abc123',  // Optional
}

Anthropic

provider: {
  type: 'anthropic',
  apiKey: process.env.ANTHROPIC_API_KEY!,
}

Azure OpenAI

provider: {
  type: 'azure-openai',
  apiKey: process.env.AZURE_OPENAI_KEY!,
  endpoint: 'https://your-resource.openai.azure.com',
  deploymentId: 'gpt-4',
}

Sampling Configuration

Basic Sampling

sampling: {
  rate: 100,  // 1 in 100 requests get L2 attestation
}

Policy Checks

sampling: {
  rate: 100,
  policies: [
    'toxicity',      // Harmful content detection
    'pii',           // PII detection
    'bias',          // Bias detection
    'prompt_injection', // Prompt injection detection
  ],
}

Adaptive Rules

sampling: {
  rate: 100,
  rules: [
    // Sample all GPT-4 requests
    {
      condition: { field: 'model', operator: 'eq', value: 'gpt-4' },
      rate: 1,
    },
    // Sample more for long prompts
    {
      condition: { field: 'content_length', operator: 'gt', value: 2000 },
      rate: 10,
    },
    // Sample specific users
    {
      condition: { field: 'user', operator: 'eq', value: 'admin' },
      rate: 1,
    },
  ],
}

Logging Configuration

logging: {
  level: 'info',         // Minimum log level
  format: 'json',        // Output format
  includeRequestId: true, // Include request ID in logs
  redactSecrets: true,   // Redact API keys in logs
}

Advanced Options

Request Batching

batching: {
  enabled: true,
  maxSize: 100,      // Max attestations per batch
  maxWaitMs: 1000,   // Max wait before flush
}

Retry Configuration

retry: {
  maxAttempts: 3,
  initialDelayMs: 100,
  maxDelayMs: 5000,
  backoffMultiplier: 2,
}

Custom Headers

headers: {
  'X-Custom-Header': 'value',
}

Validation

The sidecar validates configuration at startup:

Terminal window
npx glacis-sidecar validate

Next Steps